A SURVEY OF MACHINE LEARNING MODELS FOR INFRASTRUCTURE RESILIENCE IN FINTECH APPLICATIONS

Abstract

The survey will consider how machine learning (ML) models can be used to improve infrastructure resilience in Fintech applications. With rising complexity, data volumes, and cybersecurity threats to Fintech systems, ML offers substantial basis to the maintenance of operational continuity, prevention of fraud, anomaly detection, and real-time risk management. Such abilities are essential in ensuring system consistency and flexibility in busily developing digital financial landscapes. The paper examines different types of ML models, supervised learning, unsupervised learning, reinforcement learning and their use in fraud detection, the detection of anomalies, cybersecurity and recovery of disasters. The deployment architectures like cloud-native systems and federated learning are also discussed in terms of the advantages in scale and data secrecy. Along with this, the unification of ML with innovation-based technologies, such as blockchain and generative artificial intelligence (AI), indicates the prominence of the key indicators of system integrity, digital trust, and financial inclusion. An adequate literature research signifies the availability of the research on such topics as banking efficiency, decentralized protocols and AI-driven digital transformation. The paper also provides a comparison of strengths, challenges and future potential of ML-driven solutions in Fintech infrastructure. The survey sums up, stating that explainable AI, adaptive learning frameworks, and cross-border compliance mechanisms are needed to strengthen intelligent, resilient, and inclusive financial systems.